There are four polls in the Sunday papers – Lord Ashcroft, Opinium, Survation and YouGov.

The majority of the fieldwork for the YouGov/Sunday Times and Opinium/Observer polls was conducted before the results of the Clacton by-election were known. Opinium have topline figures of CON 28%(-4), LAB 35%(+1), LDEM 9%(+2), UKIP 17%(n/c). YouGov have topline figures of CON 32%, LAB 34%, LDEM 9%, UKIP 16%. Lord Ashcroft didn’t put voting intention figures in the headline results but they were asked and are in the tables themselves, results were CON 31%, LAB 34%, LDEM 8%, UKIP 18% – note that the poll was conducted online, whereas Ashcroft’s regular VI polls are done by phone.

Survation’s poll in the Mail on Sunday was conducted wholly on Friday – in the full glare of the post-Clacton by-election media coverage, and it shows the big boost for UKIP that we expected. Topline figures of CON 31%(nc), LAB 31%(-4), LDEM 7%(-1), UKIP 25%(+6). 25% is the highest that UKIP have ever hit in a poll. To put that in context Survation do tend to show the highest levels of support for UKIP anyway, but their previous highest was 22% and their average this year for UKIP has been 19%, so this is a significant boost anyway.

Lord Ashcroft’s poll was largely a repeat of some pre-conference questions to see if there was any impact. The survey included an opened ended question asking if people had noticed the Conservatives and Labour saying or doing anything over the last few weeks – for both parties 20% of people said that there had been a conference, the most recalled specifics were 15% remembering proposed tax cuts from the Tories, and 15% remembering that Ed Miliband forgot part of his speech. Changes in other questions were largely just a reinforcement of the existing pattern – so Labour increased their lead on “being on the side of people like me”, the Conservatives increased their lead on things like “being clear what they stand for”, “cutting the deficit” and “reforming welfare”.

Ashcroft did ask one interesting new question – a forced choice asking if people wanted Labour & Miliband to win, Labour despite Miliband, Miliband despite Labour, and the equivalent options for the Conservatives. The balance of opinion was 54% Conservative/Cameron and 46% Labour/Miliband, but the splits were interesting. Amongst Tory voters 75% wanted to see Cameron & the Conservatives win the election. Amongst Labour voters only 37% were happy with Labour and Miliband, 47% said they wanted Labour in government, even if it meant Miliband as PM. Amongst Liberal Democrat voters 66% opted for the Conservatives/Cameron, but mostly because they’d rather Cameron remained PM even if it meant the Conservatives in power. Amongst UKIP voters 66% opted for Conservatives/Cameron, 30% saying they’d want to keep Cameron even if meant the Tories, 26% because they’d rather keep the Tories even if meant Cameron.

Both YouGov and Ashcroft asked about likelihood to change at the next election. YouGov found the UKIP vote was a little softer than the Conservative and Labour parties’ – 58% of Conservatives say they are certain to vote that way, 59% of Labour voters, but only 46% of UKIP voters. Looking at potential changes, 24% of current Conservative voters said they’d consider voting UKIP at the general election, as would 11% of current Labour voters. Looking at possible movement in the opposite direction, 27% of current UKIP voters say they’d consider voting Conservative come the election, 9% would consider Labour. If the UKIP vote does get squeezed closer to the election it will benefit the Tories… but if they don’t, if the rollercoaster keeps going, there is more risk for the Tories too. Ashcroft asked the question slightly differently, only targeting those who said they might change and asking a “tick all that apply”, but the pattern of support was exactly the same.

YouGov also asked people how they might vote if their own seat looked like it would be a race between certain parties – the aim was to try and see how UKIP voters might be squeezed in marginals, but of course it also gives us some interesting pointers of possible tactical behaviour in other seats. In Conservative/Labour marginals 34% of UKIP voters say they would vote for the Conservative party, in Conservative/Liberal Democrat marginals 30% of Ukippers would vote Conservative. Imagining themselves in a Con/Lab marginal the remaining Lib Dems voters are as likely to vote tactically for the Conservatives as for Labour. 39% of Conservative voters say they’d vote tactically for the Lib Dems in Lab/LD seats, 36% of Lab voters would still be prepared to hold their nose and vote tactically for the Lib Dems in Con/LD marginals.

Tabs are here: YouGov, Survation, Ashcroft.


370 Responses to “Sunday polling round up”

1 6 7 8
  1. The polls seem to be averaging a Labour lead, still, of about 3-4% at the moment. After the recent upheaval, still a slippage of only about 1 & bit per cent since January 1. Less than 6 months to go until the general election campaign starts. I have tended to assume that the Tories would have a small lead over Labour in national vote share come May 2015, but I’m not so sure now. A canvass in a key marginal near to my house at the weekend revealed less than brilliant Tory support, too (though one voter told my colleague to f*** off!), as it happens.

  2. Barnaby Marder

    Nice to see that the electorate in England is also becoming politically engaged, and making incisive comments to canvassers. :-)

  3. Ashcroft polling:

    Share of “definitely vote that way” (n=314)

    Con 27
    Lab 36
    LD 6
    SNP 5
    UKIP 21
    Green 2

    Share of “might vote differently” (n=218)

    Con 28
    Lab 26
    LD 10
    SNP 6
    UKIP 17
    Green 9

    So Lab have a 9% lead amongst those whose decision is already made. Cons have the largest share of people who might change their mind (and not vote Conservative). UKIP support seems quite solid. And the vast bulk of Green support is from Greens who might change their mind.

  4. So today we have 3 polls and lab=cons

    Populus 71
    ICM 66
    Ashcroft 60.

    We know ICM is actually a forecast based on current VI plus adjustments for past behavior but that lifts LDs usually.

    It may be that the lower the combined Lab+Cons the better for Lab under current FPTP and with a hiigher core vote (Inc firm 2010 LDs) but not good really!

  5. @Old Nat

    Any idea when ‘normal service will be resumed and we can start to take SNP VI figures seriously again (i.e. when will they drop down again to 50% above where they were in 2010)?

    @PH

    SNP at either 5 or 6% wipes out at least 20 Labour seats (possibly 25). The present assumptions that a certain %age lead for Labour GB-wide will automatically lead to a Labour victory are, IMO highly suspect.

  6. this is arrant madness…which 20 labour seats are “wiped” out by the so-called SNP surge?

  7. Phil H and JimJam

    Well done both for imaginative ‘ casting of the runes ‘

    Lab will win the Ugly contest in May of next year, but not by much of an OM , maybe less than ten.
    (That’s my highly scientific rune casting by the way! )

  8. @chouenlai

    I have been thinking about making this point for a while now as I believe this ‘liars’ meme has got out of hand.

    I think the curse is actually something different, worse than mere lying, and I think it does comes back to a popular misapprehension of the implications of the Nick Clegg/LD tuition fees pledge.

    People think that the lesson here is that Nick and his colleagues are liars. But that isn’t the problem. What this issue has done instead is demonstrate that politicians working within the current systems (ie. Westminster/ the EU/ globalisation) may start out with noble beliefs and intentions, but on entering the world of real government, immediately find that these are dashed against the rocks of reality, affordability and (even) possibility. So we now all (wrongly) accept that it would be impossible for a government to axe tuition fees.

    It suits proponents of ‘small government’ to encourage a view that the state can’t change anything, and the Lib Dems have been willing dupes in this regard, and UKIP/SNP the main beneficiaries – essentially arguing that we need a simpler system (i.e. no EU, no rUK) within which to act.

    EM was on the brink of making this point in his conference speech but failed/forgot(!) to follow through. It’s his biggest problem – people thinking that he will not be able to enact policies based on good intentions, because of a Westminster malaise caused by external forces.

  9. John B

    I’m not sure that they ever will.

    Not that I’m saying these Scottish crossbreaks are in any way accurate, just that the pollsters (and/or their clients) are still locked into a GB mindset, that assumes that including all of the GB in a poll is in any way meaningful.

  10. Now that there have been televised leader debates, I don’t think for there can be any going back. Any leader refusing to appear would just seem “frit” and would never hear the end of it.

  11. My prediction for the GE is that Labour will get more than 29%, the Tories less than 36% and the Lib Dems less than 18%.

  12. Peter Crawford

    To be fair, the seats won’t be wiped out – just the Labour MPs.

    The Scotland Votes electoral calculator suggests that omn a Scottish vote share of SNP 39% : Lab 32% : Con 16% : LD 6% the lost Lab seats would be Aberdeen North; Aberdeen South; Ayrshire Central; Ayrshire North & Arran; Dundee West; East Kilbride, Strathaven and Lesmahagow; East Lothian; Edinburgh East; Edinburgh N & Leith; Edinburgh South; Falkirk; Kilmarnock & Loudon; Lanark & Hamilton East; Linlithgow & East Falkirk; Livingstone; Midlothian; Ochil & South Perthshire; and Stirling.

    Of course, that’s just an electoral calculator’s assumptions based on a possible (but not unlikely, voting pattern).

  13. @Oldnat

    “Sorry to disappoint – SNP is only 62% in Ashcroft.”

    LoL! I thought you were joking. Of course, 1% for PC. However the samples are somewhat on the questionable side. :-p

    However, have a piccy of the poll and the ‘Scotland Votes’ results, for posterity:

    http://www.statgeek.co.uk/wp-content/uploads/2014/10/posterity.png

    :))

  14. New thread

  15. I think Cameron ducking out of the debates could potentially be quite damaging for DC – at the moment he polls well on leadership & looking Prime Ministerial. If the public perceive him as ‘weaselling out’ then he’ll seem a lot less dignified and statesmanlike. Of course the important bit of that last sentence is ‘if the public perceive’, which depends on Pressman et al, so I won’t hold my breath.

    Also – the fact that most people don’t watch the debates doesn’t necessarily imply that they aren’t in favour of them happening.

  16. @Peter Crawford

    Put the Scottish crossbreaks into Westminster Votes Scotland (I would advise you reduce the SNP to 35% and increase Labour to 35%) and see what comes out.

    The current Scottish crossbreaks are, of course, very questionable (which is why I advise that SNP % be reduced and Labour % increased when you put the figures into the swing machine) but what no-one can deny is that since the referendum all polling companies are showing SNP support markedly increased on 2010 levels.

    Up to SNP levels of 35% in Scotland Labour lose only a few and most SNP gains are from LDs. Over 37% and suddenly huge swathes of ‘safe’ Labour territory start to fall to the SNP.

    Unless the SNP cross breaks return soon to below 35% for Westminster (as opposed to Holyrood – where SNP look as though they may have yet another overall majority in 2016 – something which was meant to be impossible for any party, remember?), Labour will be worried that there may have been a major shift in the Scottish VI patterns following the referendum.

    The UNS projection simply does not allow for this.

  17. And of course, the 5% or 6% level is on a UK basis, not a Scotland basis. This, perhaps, is where people are missing the point.

  18. @Statgeek

    Some picture! However, you’re not counting personal vote. Charlie will still be back for Skye, Wester Ross and Lochaber even if the LDs were on 0%. But few others can count on such support.

  19. @John B

    For all we know, there’s 80+ percent support for the SNP in SNP strongholds. That’s the limitations of simple % calcs.

    I believe that the SNP will make gains, but am unwilling to predict any number. More than 1. Less than 53. :))

  20. It’s not ‘cherry picking’ to observe that the Greens have been getting close to the Lib Dems – here’s the poll movement between the two with YouGov over the last six months: http://blogs.lse.ac.uk/politicsandpolicy/files/2014/10/Screen-Shot-2014-10-07-at-17.44.54.png

1 6 7 8